• 沒有找到結果。

新鮮枝葉掉落物對地表無脊椎動物群集的影響

N/A
N/A
Protected

Academic year: 2021

Share "新鮮枝葉掉落物對地表無脊椎動物群集的影響"

Copied!
41
0
0

加載中.... (立即查看全文)

全文

(1)國立臺灣師範大學生命科學系碩士論文. 新鮮枝葉掉落物對地表無脊椎動物群集 的影響 The Effects of Greenfalls on Ground-dwelling Arthropod Community. 研 究 生:陳 建 龍 Chien-Lung Chen 指導教授:李 佩 珍 博士 Pei-Jen L. Shaner 中 華 民 國 105 年 7 月.

(2)

(3) 誌謝 第一個要感謝的是我的指導教授李佩珍,何其幸運能在人生的這 個階段進入實驗室,接受老師的指導。老師在科學方法及問題解決的 能力及做法上,都給了我一個很好的模範,如何從多方角度去思考一 個問題,並且重整自己的想法脈絡,進而有條理的口述出來,謝謝老 師在這幾年的時間給的訓練,使我潛移默化的獲得成長。在老師身上 我也看到老師對生態研究的熱愛,在實際野外操作過程中會遇到很多 困難,該如何解決如何妥協,以達到順利完成研究的目的。實際參與 過後才能體會,生態學的野外研究遠比我過去所想的危險與艱辛,而 這些經歷都會成為我生命中難以忘懷的一部分,成為我的養分跟著我 繼續成長。 感謝林登秋博士與何傳愷博士擔任我的口試委員,對於研究內容 及寫作都提出了許多建議,讓這篇論文,這個研究更有意義。也謝謝 林登秋老師在研究過程中對於實驗方法的建議與討論,使這個研究更 加精確,也謝謝在各個場合聽我報告過這個研究的老師與朋友們,你 們的建議與評語都幫助我以不同的角度來檢視這個研究,使其更加完 善。 謝謝實驗室的大家,記得那些一起在湧泉池崩潰迷路,一起在煙 聲撞鬼,在草地被芒草刮的遍體鱗傷的開收籠,都是我對野外實驗不.

(4) 可抹滅的回憶。還有在實驗室揉麵團食物,大家一起在實驗室搭伙吃 晚餐,一起崩潰統計和忘記自己看過的 paper 到底寫了些什麼,都是 我在這研究生活中能繼續下去的動力,也只有你們能和我體驗到生態 研究中相同的愉悅與痛苦。謝謝伶樺學姊在研究過程中的各種幫助, 有困難的時候第一個想到的就是和你討論。謝謝佩真,一樣是做跟老 鼠不相干的研究,一起打球,繼承讀很久的座位,有一些有的沒的生 活瑣事都可以找你聊天。謝謝諠憶、苡柔、艾芸、慶賀、鴻鈞、家綉、 偉廷,這個研究的過程都有大家給予的點點滴滴,和你們一起的實驗 室生活讓我感覺像個家。最後謝謝曾經在我野外研究提供幫助的允瑄、 立辰、仲康、騏銘、立中、子維、士凱,以及武陵工作站的各位工作 人員,沒有你們的幫助,可能實驗就會增添許多變數。 最後謝謝我的家人,研究所這幾年的經濟支持,以及心理上的信 賴與付出,都是讓我能安心完成碩士學位的重要因素,如果這篇研究 能對知識有著這麼一點點的貢獻,那麼你們也是功不可沒。謝謝沅道 在我接近收尾的這段時間,忍受我的崩潰暴躁與無理取鬧。謝謝我身 邊的所有朋友們,你們的一句鼓勵都是我能走到今天的動力,也謝謝 我未能在這裡提到的朋友們,希望這個研究的成果也能讓你們感到與 有榮焉。.

(5) Table of Contents. Abstract……………….………..………………………………..………1 摘要………………………………………………………………………2 Introduction……………………...………………………………………4 Materials and Methods….…...........…………………….……………7 Study site………………………..….……………….……..………7 Greenfall treatment ...……….…….……………….……….…...…7 Arthropod sampling…………….…………………............................8 Data analysis…………………………...………………………..…...9 Results…………………………………………………...……………...10 Discussion…………………………...………………………...………..12 Reference……………………………...………………………..………13 Tables & Figures…………………………………………………......18 Supplementary Materials………………………………...……………30.

(6) Abstract Resource pulses are rare, brief and intense episodes of increased resource availability in space and time. Greenfalls caused by typhoons is a resource pulse event for ground-dwelling arthropods. In this study, I artificially created a resource pulse of greenfalls in a Pinus taiwanensis plantation to simulate one that is caused by typhoons in a subtropic montane forest in central Taiwan. I set up 20 plots (5 m x 5 m) at the study site, of which 10 were randomly assigned to receive 10 kg of fresh P. taiwanensis greenfalls in the summer of 2013 whereas the other 10 plots were left untouched as controls. I monitored the activity-density and biomass of ground-dwelling arthropods using the pitfall traps for 8 months (2 months of pre-treatment time period and 6 months of post-treatment time period). The results suggest that the greenfalls did not have effects on arthropod active-density and biomass as a whole nor did it affect the predators and detritivores. However, I detected positive, bottom-up effects from the greenfalls to the intermediate consumers (most likely herbivores), both in terms of their actual and proportional biomass. This study demonstrated potential bottom-up effects of the greenfalls on the above-ground consumers, which did not propagate to the top predators. Furthermore, the greenfalls likely have limited influences on the detritivores. The strong seasonal fluctuations in arthropod activity-density and biomass at this subtropical montane forest may have limited the influences of the greenfalls. Key words: natural disturbance, trophic interaction, typhoon, food web, resource fluctuation, resource—consumer dynamics 1.

(7) 摘要 間歇性資源是指短暫、罕見並且高強度的可利用性資源在時間與空間 的尺度下增加的事件。對地表的無脊椎動物來說,颱風所造成的新鮮 枝葉掉落物就是一種間歇性資源的事件。在這個研究中,我在以臺灣 二葉松為主的人工林中,人為製造了一個模擬颱風經過臺灣中部亞熱 帶山區森林所產生新鮮枝葉掉落物的間歇性資源事件。我建立了二十 個五公尺乘以五公尺的網格,隨機指定其中十個於 2013 年七月接受 十公斤的新鮮二葉松枝葉做為處理組,另外十個則不進行處理,以做 為控制組。我利用掉落式陷阱採集地表無脊椎動物樣本做為活動密度 與生物量的指標,共採集八個月,其中處理前兩個月,處理後六個月。 結果顯示,新鮮枝葉掉落物的處理並沒有對無脊椎動物群集的活動密 度與生物量有整體上的影響,對於掠食者與碎屑食者的活動密度與生 物量也無影響。然而我發現新鮮枝葉掉落物對於中間消費者有一個由 營養階層下方往上傳遞的正向影響,而在中間消費者當中,經由新鮮 枝葉掉落物而受益的類群,較可能為植食性動物。上述這個正向的影 響同時表現在中間消費者的生物量分析以及其生物量占所有類群總 合的比例分析上。這個研究發現了新鮮枝葉掉落物對於地表的消費者 有營養階層由下往上的效應,但是這個效應並沒有傳到最上層的掠食 者。此外,新鮮枝葉掉落物對於碎屑食性的消費者似乎沒有太大的影 2.

(8) 響,由於其生活史導致活動密度與生物量有強烈的季節性變動,也限 制了新鮮枝葉掉落物對碎屑食性消費者的影響。. 關鍵字:自然擾動,營養階層的交互作用,颱風,食物網,資源波動, 資源與消費者之間的動態. 3.

(9) Introduction Resource pulses, episodes of increased resource availability in space and time (Yang et al. 2008), can affect species interactions, food web structure, and energy flow in an ecosystem (e.g. Jedrzejewska and Jedrzejewski 1998; Haddad et al. 2000; McShea 2000; Schmidt and Ostfeld 2003; Stapp and Polis 2003; Schmidt and Ostfeld 2008; also see the meta-analysis by Yang et al. 2010). Many different types of resource pulse events have been recognized across terrestrial ecosystems, including mast production of flowers, fruits or seeds (Kelly 1994; Ostfeld et al. 1996; Curran and Leighton 2000; McShea 2000), insect outbreaks or aggregations (Carlton and Goldman 1984; Haney 1999; Yang 2004), large inputs of animal carcasses, dung, or urine (Rose and Polis 1998; Peek and Forseth 2003; Wilmers et al. 2003; Yang 2006), and typhoon-triggered greenfalls (Lodge et al. 1994). Different resource pulse events may have different characteristics (e.g. magnitudes, lengths, frequencies) and influence different consumers (e.g. microbes, plants, invertebrates, vertebrates). Therefore, different resource pulse events are likely to produce different ecological outcomes. Typhoons can cause substantial physical damages to trees (Whigham et al. 1991; Boose et al. 1994) and thereby create resource pulses of greenfalls (Lodge et al. 1994; Yang et al. 2008). Greenfalls are composed mainly of fresh or immature leaves, and hence are nutrient-rich plant resources. This is in sharp contrast to the aging leaves from which the nutrients had been translocated to other organs prior to falling (Whigham et al. 1991; Lodge et al. 1991). Greenfalls generated by hurricanes in a 4.

(10) montane rain forest was shown to create an influx of nutrients to the forest ground that was more than twice the annual average (hurricane-induced N and P inputs: 4.2 g m-2, 0.15 g m-2; annual average N and P inputs: 1.9 g m-2, 0.06 g m-2; Lodge et al. 1991). Furthermore, the quantity and quality of greenfalls generated by typhoons may be different in different types of forests. For example, Wang et al. (2013) compared litterfall dynamics between a natural hardwood forest and a Chinese-fir plantation, both of which experienced the same four typhoons in 2008. The litterfall generated by the four typhoons contributed 22% more to the annual litterfall in the Chinese-fir plantation (81%, or 6000 out of 7400 kg/ha) than in the natural hardwood forest (59%, or 6630 out of 11400 kg/ha). This study suggests that typhoons may have relatively larger impacts on nutrient dynamics in the fir plantations than the natural hardwood forests. Arthropods are sensitive to changes in microenvironments, making them suitable indicator species for environmental perturbation (Kremen 1992; Kremen et al. 1993; Lieberman and Dock 1982), such as typhoon-triggered greenfalls. Furthermore, an arthropod community typically comprises many tightly-linked trophic groups, such as detritivores, herbivores, omnivores, predators and parasitoids, and each of them may respond differentially to a single resource event (e.g. Hoekman et al. 2011; Murphy et al. 2012). The sensitivity of arthropods to environmental perturbation, and the dynamics of their trophic interactions, make them an excellent system to study the impacts of greenfalls on consumer communities. 5.

(11) Resource pulses may involve both bottom-up and top-down processes, and the outcomes often depend on the temporal scale of these dynamics. Specifically, a nutrient pulse may increase plant productivity at time 0, herbivore abundance at time 1, and predator abundance at time 2. If the nutrient pulse diminishes by time 1, the herbivores benefit. On the other hand, if the nutrient pulse lasts through time 2, the predators may also benefit, leading to an initial increase in herbivores only followed by a subsequent decrease due to predation. For example, Murphy et al. (2012) added a nutrient pulse to a spartina (Spartina alterniflora) system, which led to increased plant biomass and leaf N content, followed by increased abundances of herbivorous and predatory insects. They detected top down effects in the following year after the treatment. It is important to track a resource pulse event for a sufficient amount of time and at an appropriate frequency in order to capture the temporal dynamics in consumer response. This study was conducted at a pine plantation of Pinus taiwanensis in the Shei-Pa National Park in Taiwan. I artificially simulated a greenfall event that could have been generated by a typhoon to investigate arthropod responses over a time period of 6 months. I predicted that: (1) P. taiwanensis greenfalls trigger a bottom-up process whereby the activity-density (a proxy for abundance) and/or biomass of all arthropods combined will increase, with detritivores and herbivores showing the increase first, followed by omnivores and predators; (2) P. taiwanensis greenfalls trigger a delayed top-down process whereby herbivores and detritivores decrease in activity-density and/or biomass following the 6.

(12) increase of predators and omnivores.. Material and Methods Study site The study was conducted in an evergreen forest (P. taiwanensis) in the Shei-Pa National Park (121°18’E, 24°21’N; elevation: 1815-1945 m). The mean annual precipitation is 1100mm and mean annual temperature is 13℃ (lowest in January at 5℃ and highest in July at 19℃, 2005-2008, Taiwan Forestry Bureau). Central Taiwan experiences one typhoon per year on average (Wang et al. 2013). Greenfall treatment I set up 10 pairs of 5 m x 5 m plots, with a total of 20 plots. The distance between plots was at least 50 m, and all plots were more than 10 m away from the edge of the forest (Fig. 1). One plot in each pair was randomly assigned to receive the greenfalls (the treatment plot), while the other one served as control. Each treatment plot was given 10 kg of fresh P. taiwanensis greenfalls on July 7th, 2013 to simulate the greenfalls brought about by a typhoon, while the control plot reminded intact. The amount of the greenfalls (4000kg/ha) given was twice the amount reported in Wang et al. (2013) for the typhoon Sinlaku (instantaneous maximum wind velocity 51 m/s, precipitation 860 mm) at Lienhuachi Experimental Forest, a site comprising both natural hardwood forests and fir plantations (c. 67.5 km from current study site). On July 13th, 2013, one week after the greenfall treatment, the study site was hit by the typhoon Soulik (instantaneous maximum wind speed 38m/s, precipitation 307 mm at 7.

(13) Lishan, about 13 km from current study site, Taiwan Central Weather Bureau Typhoon DateBase). On July 18th, I removed all fresh branches from the control plots but left the treatment plots with the additional, naturally-occurring greenfalls. Arthropod sampling The arthropods were sampled using pitfall traps. Within each plot, I placed three pitfall traps, with a minimum distance of 1 m between traps. The pitfall trap was made of a 50-ml centrifuge tube (11.5 cm in depth, 3 cm in diameter). A small plastic cover (8 cm in diameter) was placed 5-6 cm above each trap to prevent rain water from falling in. Each trap was filled with 25 ml (about 6 cm in depth) of 4% formaldehyde solution and a few drops of unscented dish detergent as an interface agent to reduce the surface tension (Tulp and Schekkerman 2008). The trap was buried at a depth such that the top of the tube was flush with the surface of the ground. I set the traps on May 12th, 2013, and collected samples every 15±3 days until December 27th. I applied the greenfall treatment on July 7th. Therefore, I had 4 pre-treatment samples and 11 post-treatment samples, with a total of 15 sampling periods. To remove soils, I gently rinsed the samples with clean water in a mesh sieve (0.420 m/m opening). The remnants were stored at 75% ethanol for taxonomic identification. I classified the arthropods to order or family level based on their morphology under a dissecting microscope (10x ocular lens with 0.8x-5x objective lens). I counted the number of individuals in each taxonomic group. All individuals were then dried in an oven at 60℃ to constant weight for biomass estimates. 8.

(14) Data analysis The plot is the experimental unit. The numbers and biomass of arthropods were averaged across the three pitfall traps to obtain plot-level responses. I analyzed the arthropod data at three ecological levels: community, trophic group and taxonomic group. At community level, I pooled the numbers and biomass of all arthropod individuals. At trophic-group level, I focused on top predators (i.e. Araneae, Chilopoda and Staphylinidae), intermediate consumers (i.e. herbivores including Belidae, Gryllidae, Tetrigidae and Rhaphidophoridae, and omnivores including Carabidae, Formicidae, and Blattidae), and detritivores (i.e. Diplopoda, Collembola, Isopoda and Oligochaete). At taxonomic-group level, I looked at six most common taxa (i.e. Araneae, Staphylinidae, Carabidae, Formicidae, Collembola and Isopoda). I used generalized linear mixed models to test the effects of the greenfall treatment on the activity-density and biomass of the arthropods. The greenfall treatment, sampling time and their interaction are treated as fixed effects, with each plot repeatedly measured across sampling time periods as a random effect. To reduce the amount of temporal variation from one sampling period to the next, I condensed the 15 sampling periods to 7 by taking the averages of 2-3 consecutive samples (2 consecutive samples between May 27th and November 13th, and 3 consecutive samples between November 29th and December 27th). In addition to actual activity-density and biomass values, I also used the ratios of the activity-density and biomass after the treatment to that prior to the treatment as the response variable for the generalized linear mixed 9.

(15) model, which produced qualitatively similar results (Table S3&S4). To detect possible changes in community composition that may influence ecosystem processes, I calculated the proportional biomass for each of the three trophic groups (i.e. the biomass of a trophic group divided by the combined biomass of all three trophic groups in a given plot), both prior to and after the greenfall treatment (all sampling time periods pooled). I used generalized linear mixed models to test the effects of the greenfall treatment on the proportional biomass of the trophic groups. The greenfall treatment, the pre-/post-treatment time period, and their interactions, were treated as the fixed effects, with each plot repeatedly measured between the pre-/post-treatment time periods as a random effect. The generalized linear mixed models were fitted with either Poisson or negative binominal distribution with a log link function (Pearson Chi-Square / DF < 3.9 for all models). All statistical analyses were done in SAS 9.4.. Results The greenfall treatment did not appear to influence the trophic or taxonomic compositions of the arthropod community. Across the control and treatment plots during both pre- and post-treatment periods, the detritivores were the most dominant trophic group by number (87%-91%; Fig. 2a) and by biomass (65%-86%; Fig. 2b), whereas intermediate consumers and predators each represented less than 10% by number and 35% by biomass (Fig. 2). Similarly, regardless of the greenfall treatment, 10.

(16) Collembola was the most dominant taxon by number (86%-90%; Fig. 3a) and Isopoda the most dominant by biomass (60%-83%; Fig. 3b), whereas other commonly seen taxa each represented less than 5% by number of 25% by biomass (Fig. 3). The proportional biomass of the predators remained unchanged before and after the treatment (Table 1). The proportional biomass of the intermediate consumers increased in the treatment plots, but not in the control plots, after the greenfall addition. Corresponding to the increase in proportional biomass of the intermediate consumers, the detritivores decreased in proportional biomass in the treatment plots after the greenfall addition (Fig 4 a, b&c). The greenfall treatment did not affect either the activity-density or biomass of all arthropods combined (Table 2). The greenfall treatment did not affect either the activity-density or the biomass of the predators. Similarly, it did not affect the active-density of the intermediate consumers or the biomass of the detritivores (Table 2). The treatment plots had a higher biomass of the intermediate consumers between July 24th and August 7th and between October 25th and November 13th (Fig. 5f). On the other hand, the activity-density of the detritivores was higher in the control plots than in the treatment plots between May 27th and June 10th, prior to the treatment (Fig. 5g). Among the six taxonomic groups, only the active-density of Collembola reacted to the greenfall treatment (Table 3). The active-density of Collembola in the control plots was higher than in the treatment plots between May 27th and June 10th, prior to the treatment 11.

(17) (Fig. 6i).. Discussion The actual and proportional biomass of the intermediate consumers showed a rapid and positive response to the greenfalls. This pattern suggests that the greenfalls may trigger a bottom-up process that benefits some intermediate consumers (most likely herbivorous consumers rather than omnivorous consumers such as Carabidae and Formicidae). However, I did not detect a delayed, top-down effect through increased predation on the intermediate consumers or detritivores. Although the active-density of Collembola suggests potential suppressing effects of the greenfall treatment, the difference occurred prior to the treatment. Therefore, it also could be due to pre-existing environmental variation unrelated to the greenfall treatment. Neither of the two intermediate-consumer taxa, Carabidae and Formicidae, responded to the greenfalls. However, the intermediate consumer as a group, responded positively to the greenfalls. Therefore, it is likely that the herbivores, rather than the omnivores (ground beetles and ants are omnivores), within the intermediate consumer group, drove the positive response. The activity-density and biomass of the arthropods in current study exhibited strong temporal fluctuations, which is not uncommon for invertebrate communities. However, it does limit my ability to detect the treatment effect. With a higher sample size, future studies may be able to reveal more complex responses at finer taxonomic level. Nevertheless, 12.

(18) the time frame of this study should be sufficient to capture some of the temporal dynamics in the arthropod community following a typhoon-induced greenfall event. Towards the end of the study in December, both the means and variations in arthropod activity-density and biomass had become very low, indicating that many arthropods were no longer active. The added greenfalls were visually undetectable towards the end of the study. However, it is possible that they could have inter-annual effects on the arthropods as they enter into soil nutrient pools. This study offers one of the few empirical cases on how animal consumers respond to a greenfall-pulse event. Depending on the type of the forests and the characteristics of the typhoons, greenfalls may comprise branches and twigs of varying sizes and nutrient quality. Consequently, they can have dynamic impacts on biological communities as a food resource, as well as an abiotic force that alters habitat structure and microclimates. Yet our understanding of greenfalls as a pulse resource event remains rudimentary. The intensity and frequency of typhoons are predicted to increase under global warming (Chan and Liu 2004). Therefore, this study is a timely contribution to our understanding of typhoon-induced ecological dynamics.. Reference Boose ER, Foster DR, Marcheterre F (1994) Hurricane impacts to tropical and temperate forest landscapes. Ecological Monographs 64: 369–400. 13.

(19) Carlton RG, Goldman CR (1984) Effects of a massive swarm of ants on ammonium concentrations in a subalpine lake. Hydrobiologia 111: 113–117. Chan JCL, Liu KS (2004) Global warming and western north pacific typhoon activity from an observational perspective. Journal of Climate: 17: 4590–4602. Curran LM, Leighton M (2000) Vertebrate responses to spatiotemporal variation in seed production of mast-fruiting Dipterocarpaceae. Ecological Monographs 70: 101–128. Haddad NM, Haarstad J, Tilman D (2000) The effects of long-term nitrogen loading on grassland insect communities. Oecologia 124: 73–84. Haney JC (1999) Numerical response of birds to an eruption of elm spanworm (Ennomos subsignarius; Geometridae: Lepidoptera) in old-growth forest of the Appalachian Plateau, USA. Forest Ecology and Management 120: 203–217. Hoekman D, Dreyer J, Jackson RD, Townsend PA, Gratton C (2011) Lake to land subsidies: experimental addition of aquatic insects increases terrestrial arthropod densities. Ecology 92: 2063–2072. Jedrzejewska B, Jedrzejewski W (1998) Predation in vertebrate communities: the Bialowieza primeval forest as a case study. Ecological Studies 135:450. Springer– Verlag, Berlin – Heidelberg – New York. Kelly D (1994) The evolutionary ecology of mast seeding. Trends in Ecology and Evolution 9: 465–471. 14.

(20) Kremen C (1992) Assessing the indicator properties of species assemblages for natural areas monitoring. Ecological Applications 2: 203–217. Kremen C, Colwell RK, Erwin TL, Murphy DD, Noss RF, Sanjayan MA (1993) Terrestrial arthropod assemblages: their use in conservation planning. Conservation Biology 7: 796–808. Lieberman S, Dock CF (1982) Analysis of leaf litter arthropod fauna of a lowland tropical evergreen forest site (La Selva, Costa Rica). Revista de Biologia Tropical, 30: 27-34. Lodge DJ, McDowell WH, McSwiney CP (1994) The importance of nutrient pulses in tropical forests. Trends in Ecology and Evolution 9: 384–387. Lodge DJ, Scatena FN, Asbury CE, Sanchez MJ (1991) Fine litterfall and related nutrient inputs resulting from hurricane Hugo in subtropical wet and lower montane rain forests of Puerto Rico. Biotropica 23: 336–342. McShea WJ (2000) The influence of acorn crops on annual variation in rodent and bird populations. Ecology 81: 228–238. Murphy SM, Wimp GM, Lewis D, Denno RF (2012) Nutrient presses and pulses differentially impact plants, herbivores, detritivores and their natural enemies. PLoS ONE 7: e43929. Ostfeld RS, Jones CG, and Wolf JO (1996) Of mice and mast. BioScience 46: 323–330. Peek MS, Forseth IN (2003) Enhancement of photosynthesis and growth of an aridland perennial in response to soil nitrogen pulses generated 15.

(21) by mule deer. Environmental and Experimental Botany 49: 169-180. Rose MD, Polis GA (1998) The distribution and abundance of coyotes: the effects of allochthonous food subsidies from the sea. Ecology 79: 998–1007. Schmidt KA, Ostfeld RS (2003) Songbird populations in fluctuating environments: predator responses to pulsed resources. Ecology 84: 406–415. Schmidt KA, Ostfeld RS(2008)Numerical and behavioral effects within a pulse-driven system: consequences for shared prey. Ecology 89: 635–646. Stapp P, Polis GA (2003) Influence of pulsed resources and marine subsidies on insular rodent populations. Oikos 102: 111–123. Taiwan Central Weather Bureau. Typhoon DateBase: http://rdc28.cwb.gov.tw/TDB/ntdb, data accessed on June 24th, 2016. Taiwan Forestry Bureau. http://www.forest.gov.tw/lp.asp?CtNode=1996&CtUnit=798&Base DSD=7&mp=350, data accessed on June 24th, 2016. Tulp I, Schekkerman H (2008) Has prey availability for arctic birds advanced with climate change? Hindcasting the abundance of tundra arthropods using weather and seasonal variation. Arctic 61: 48–60. Wang HC, Wang SF, Lin KC, Shaner PL, Lin TC (2013) Litterfall and element fluxes in a natural hardwood forest and a Chinese-fir plantation experiencing frequent typhoon disturbance in central Taiwan. Biotropica 45: 541–548. Whigham DF, Olmsted I, Cano EC, Harmon ME (1991) The impact of 16.

(22) hurricane Gilbert on trees, litterfall, and woody debris in a dry tropical forest in the northeastern Yucatan Peninsula. Biotropica 23: 434–441. Wilmers CC, Stahler DR, Crabtree RL, Smith DW, Getz WM (2003) Resource dispersion and consumer dominance: scavenging at wolfand hunter-killed carcasses in Greater Yellowstone, USA. Ecology Letters 6: 996–1003. Yang LH (2004) Periodical cicadas as resource pulses in North American forests. Science 306: 1565–1567. Yang LH (2006) Interactions between a detrital resource pulse and a detritivore community. Oecologia 147: 522–532. Yang LH, Bastow JL, Spence KO, Wright AN (2008) What can we learn from resource pulses? Ecology 89: 621–634. Yang LH, Edwards KF, Byrnes JE, Bastow JL, Wright AN, Spence KO (2010) A Meta-analysis of resource pulse – consumer interactions. Ecological Monographs 80: 125–151.. 17.

(23) Tables & Figures Table 1. Generalized linear mixed models of the ratio of functional groups biomass between before and after treatment. Significant effects in bold. Effect. Num DF. Den DF. F. P. Greenfall. 1. 18. 0.32. 0.58. Time. 1. 18. 1.76. 0.20. Greenfall x time. 1. 18. 0.00. 0.96. Predators. Intermediate consumers (omnivores and herbivores) Greenfall. 1. 18. 4.62. 0.05. Time. 1. 18. 0.81. 0.38. Greenfall x time. 1. 18. 68.26. <.0001. Greenfall. 1. 18. 0.87. 0.36. Time. 1. 18. 0.11. 0.74. Greenfall x time. 1. 18. 6.83. 0.02. Detritivores. 18.

(24) Table 2. Generalized linear mixed models of the activity-density and biomass of all arthropods combined and of different trophic groups. Significant effects in bold. Effect. Num. Den. Activity-density. Biomass (mg). DF. DF. F. P. F. P. Greenfall. 1. 18. 0.40. 0.54. 1.17. 0.29. Time. 6. 108. 8.86. <0.0001. 8.58. <0.0001. Greenfall x time. 6. 108. 1.93. 0.08. 2.06. 0.06. Greenfall. 1. 18. 0.65. 0.43. 0.47. 0.50. Time. 6. 108. 10.07. <.0001. 4.76. 0.0002. Greenfall x time. 6. 108. 1.08. 0.38. 0.76. 0.60. All arthropods. Predators. Intermediate consumers (omnivores and herbivores) Greenfall. 1. 18. 1.05. 0.32. 1.32. 0.27. Time. 6. 108. 11.48. <.0001. 20.26. <.0001. Greenfall x time. 6. 108. 0.27. 0.95. 2.5. 0.03. Greenfall. 1. 18. 0.50. 0.49. 2.96. 0.10. Time. 6. 108. 6.37. <.0001. 6.03. <.0001. Greenfall x time. 6. 108. 2.48. 0.03. 0.84. 0.54. Detritivores. 19.

(25) Table 3. Generalized linear mixed models of the activity-density and biomass of the six most dominant taxonomic groups. Significant effects in bold. Effect. Num. Den. Activity-density. Biomass (mg). DF. DF. F. P. F. P. Greenfall. 1. 18. 0.45. 0.51. 0.74. 0.40. Time. 6. 108. 37.73. <.0001. 3.58. 0.003. Greenfall x time. 6. 108. 0.64. 0.70. 0.92. 0.48. Greenfall. 1. 18. 0.49. 0.49. 0.01. 0.91. Time. 6. 108. 2.69. 0.02. 3.53. 0.003. Greenfall x time. 6. 108. 1.69. 0.13. 0.69. 0.66. Greenfall. 1. 18. 0.09. 0.77. 0.30. 0.59. Time. 6. 108. 7.00. <.0001. Greenfall x time. 6. 108. 0.23. 0.95. 0.51. 0.80. Greenfall. 1. 18. 0.54. 0.47. 1.86. 0.19. Time. 6. 108. 14.76. <.0001. Greenfall x time. 6. 108. 0.30. 0.93. 0.99. 0.43. Greenfall. 1. 18. 0.47. 0.50. 0.01. 0.92. Time. 6. 108. 4.19. 0.0008. 8.35. <.0001. Greenfall x time. 6. 108. 2.82. 0.01. 1.03. 0.41. Araneae. Staphylinidae. Carabidae. 13.93 <.0001. Formicidae. 10.82 <.0001. Collembola. 20.

(26) Isopoda Greenfall. 1. 18. 5.22. 0.03. 7.51. 0.01. Time. 6. 108. 2.62. 0.02. 3.57. 0.003. Greenfall x time. 6. 108. 0.82. 0.56. 0.71. 0.64. 21.

(27) Figure legends Fig. 1. The satellite image of the study site and the locations of the plots. The blue flags denote the locations of the control plots; the red pins denote the locations of the greenfall-treatment plots. The three locations of the Pinus taiwanensis from which I obtained the greenfalls used in this experiment are indicated by the tree symbols. Fig. 2. Relative frequencies of arthropod trophic groups in the control and greenfall-treatment plots. (a) Relative frequencies based on the number of individuals. (b) Relative frequencies based on biomass. Three trophic groups are included: detritivores (i.e. Diplopoda, Collembola, Isopoda and Oligochaete), intermediate consumers (herbivores i.e. Belidae, Gryllidae, Tetrigidae and Rhaphidophoridae, and omnivores i.e. Carabidae, Formicidae, and Blattidae), and predators (i.e. Araneae, Chilopoda and Staphylinidae). Fig. 3. Number of individuals captured and their biomass by arthropod taxonomic group in the control and greenfall-treatment plots. (a) The number of individuals captured in the control plots. (b) The number of individuals captured in greenfall-treatment plots. (c) The biomass in the control plots. (d) The biomass in the greenfall-treatment plots. Six taxonomic groups are included: Araneae (Ara), Staphylinidae (Sta), Carabidae (Car), Formicidae (For), Collembola (Col) and Isopoda (Iso). Individuals that belong to other taxonomic groups are placed into the ―other‖ group (Oth). The open bars denote the pre-treatment time period, and the filled bars the post-treatment time period. Fig. 4. The ratio of biomass by three functional groups between 22.

(28) before and after treatment. (a) Predators. (b) Intermediate consumers. (c) Detritivores. Fig. 5. Activity-density and biomass of all arthropods and of the three trophic groups. The left panel is the activity-density, and the right panel is the biomass. (a,b) All arthropods. (c,d) Predators. (e,f) Intermediate consumers. (g,h) Detritivores. The solid line denotes the control plots whereas the dash line denotes greenfall-treatment plots. The shaded box denotes the time period for which the greeenfall treatment (July 7th) and the typhoon Soulik (July 13th) occurred. The asterisks denote significant differences between the control and treatment plots for a given time period based on post-hoc comparisons. Fig. 6. Activity-density and biomass of six major taxonomic groups. The left panel is the activity-density, and the right panel is the biomass. (a,b) Araneae. (c,d) Staphylinidae. (e,f) Carabidae. (g,h) Formicidae. (i,j) Collembola. (k,l) Isopoda. The solid line denotes the control plots whereas the dash line denotes greenfall-treatment plots. The shaded box denotes the time period for which the greeenfall treatment (July 7th) and the typhoon Soulik (July 13th) occurred. The asterisks denote significant differences between the control and treatment plots for a given time period based on post-hoc comparisons.. 23.

(29) Fig. 1.. 24.

(30) Fig. 2.. 25.

(31) Fig. 3. 26.

(32) Fig. 4. 27.

(33) Fig. 5. 28.

(34) Fig. 6. 29.

(35) Supplementary Materials In order to understand whether my treatment would affect the microhabitat or not, I repeated my treatment in a smaller scale. I randomly draw five pairs of blocks from ten pairs in my experiment field, ten blocks in total. I set HOBO at the central cell (1mx1m) of these ten blocks at 17:50 of September 28 in 2014 to measure the temperature and humidity. I gave 0.4kg fresh P. taiwanensis branches which was cut off in three days to the treatment block at 17:50 of September 29 in 2014. I ended the experiment at 12:50 of October 1 in 2014. I tested the temperature and humidity of the data with generalized linear mixed model. I made treatment, before or after treatment and hours in a day as fixed factor; hours in a day as random factor. The model of temperature fit with the normal distribution, while humidity fit with the negative binominal distribution. Because of the random factor of humidity didn’t explain too much variance, I removed it from the model. The result of temperature and humidity showed that my treatment didn’t change the temperature and humidity between treatment blocks and control blocks. However, the treatment blocks were significantly drier than control blocks (F1, 584=4.49, p=0.0346), and there was no difference in temperature (F1, 392=2.29, p=0.131). That means maybe there were some systemic difference between my treatment and control blocks.. 30.

(36) Fig. S1. Temperature and humidity data of HOBO. (a,b) are temperature data, while (c,d) are humidity data. (a,c) are the data before treatment, while (b,d) are the data after treatment. The solid line means control, and the dash line means treatment.. 31.

(37) Table S1. The results of generalized linear mixed model on temperature. Variable. DF. F. P. (numerator, denominator) trt. 1, 392. 2.29. 0.131. post_trt. 1, 392. 2.18. 0.1402. trt*post_trt. 1, 392. 0.09. 0.7596. hour. 23, 192. 386.09. <.0001. trt*hour. 23, 392. 2.35. 0.0005. post_trt*hour. 23, 392. 12.9. <.0001. trt*post_trt*ho. 23, 392. 0.85. 0.6645. ur. 32.

(38) Table S2. The results of generalized linear mixed model on humidity. Variable. DF. F. P. (numerator, denominator) trt. 1, 584. 4.49. 0.0346. post_trt. 1, 584. 0.01. 0.9107. trt*post_trt. 1, 584. 0. 0.9739. hour. 23, 584. 0.78. 0.7575. trt*hour. 23, 584. 0.08. 1. post_trt*hour. 23, 584. 0.11. 1. trt*post_trt*ho. 23, 584. 0. 1. ur. 33.

(39) Table S3. Generalized linear mixed models of the activity-density ratios or the biomass ratios for all arthropods combined and for the three trophic groups. The ratios are calculated by dividing the post-treatment activity-density or biomass with the pre-treatment activity-density or biomass. Significant effects in bold. Effect. Num. Den. Activity-density. Biomass (mg). DF. DF. F. P. F. P. Greenfall. 1. 90. 2.84. 0.10. 0.10. 0.76. Time. 4. 90. 0.80. 0.53. 4.58. 0.002. Greenfall x time. 4. 90. 0.15. 0.96. 2.30. 0.07. Greenfall. 1. 90. 0.61. 0.44. 0.00. 0.95. Time. 4. 90. 4.15. 0.004. 2.23. 0.07. Greenfall x time. 4. 90. 0.42. 0.79. 0.36. 0.84. All arthropods. Predators. Intermediate consumers (omnivores and herbivores) Greenfall. 1. 90. 0.04. 0.84. 0.38. 0.55. Time. 4. 90. 3.54. 0.01. 11.15. <.0001. Greenfall x time. 4. 90. 0.12. 0.98. 6.14. 0.0002. Greenfall. 1. 90. 4.05. 0.05. 14.61. 0.0002. Time. 4. 90. 0.45. 0.77. 3.70. 0.008. Greenfall x time. 4. 90. 0.17. 0.96. 5.13. 0.0009. Detritivores. 34.

(40) Table S4. Generalized linear mixed models of the activity-density ratios or the biomass ratios for the six most dominant taxonomic groups. The ratios are calculated by dividing the post-treatment activity-density or biomass with the pre-treatment activity-density or biomass. Significant effects in bold. Effect. Num. Den. Activity-density. Biomass (mg). DF. DF. F. P. F. P. Greenfall. 1. 90. 0.18. 0.67. 0.01. 0.92. Time. 4. 90. 5.06. 0.001. 2.71. 0.03. Greenfall x time. 4. 90. 0.06. 0.99. 0.78. 0.54. Greenfall. 1. 90. 2.10. 0.15. 0.13. 0.71. Time. 4. 90. 2.20. 0.08. 2.85. 0.03. Greenfall x time. 4. 90. 1.01. 0.41. 0.19. 0.94. Greenfall. 1. 90. 0.10. 0.75. 3.00. 0.09. Time. 4. 90. 2.59. 0.04. 1.93. 0.11. Greenfall x time. 4. 90. 0.11. 0.98. 1.27. 0.29. Greenfall. 1. 90. 0.05. 0.82. 0.02. 0.88. Time. 4. 90. 1.10. 0.36. 1.34. 0.26. Greenfall x time. 4. 90. 0.07. 0.99. 0.21. 0.93. 1. 90. 5.50. 0.02. 0.38. 0.54. Araneae. Staphylinidae. Carabidae. Formicidae. Collembola Greenfall. 35.

(41) Time. 4. 90. 0.58. 0.68. 0.64. 0.64. Greenfall x time. 4. 90. 0.30. 0.88. 0.14. 0.97. Greenfall. 1. 90. 1.19. 0.28. 20.30. <.0001. Time. 4. 90. 0.90. 0.47. 5.12. 0.0009. Greenfall x time. 4. 90. 0.28. 0.89. 7.68. <.0001. Isopoda. 36.

(42)

參考文獻

相關文件

[This function is named after the electrical engineer Oliver Heaviside (1850–1925) and can be used to describe an electric current that is switched on at time t = 0.] Its graph

For ex- ample, if every element in the image has the same colour, we expect the colour constancy sampler to pro- duce a very wide spread of samples for the surface

Given a shift κ, if we want to compute the eigenvalue λ of A which is closest to κ, then we need to compute the eigenvalue δ of (11) such that |δ| is the smallest value of all of

It is my pleasure to welcome our special guest Miss Linda Chu, the chairperson of the New Life Group, and all of you to our annual Cultural Festival. The Cultural Festival is

We explicitly saw the dimensional reason for the occurrence of the magnetic catalysis on the basis of the scaling argument. However, the precise form of gap depends

Miroslav Fiedler, Praha, Algebraic connectivity of graphs, Czechoslovak Mathematical Journal 23 (98) 1973,

Microphone and 600 ohm line conduits shall be mechanically and electrically connected to receptacle boxes and electrically grounded to the audio system ground point.. Lines in

The remaining positions contain //the rest of the original array elements //the rest of the original array elements.